Identification of outliers and positive deviants for healthcare improvement: looking for high performers in hypoglycemia safety in patients with diabetes

Using Positive Deviance to Enhance HIV Care Retention in South Africa: Development of a Compassion-Focused Programme to Improve the Staff and Patient Experience

Background HIV burden remains high in South Africa despite intensive efforts to curtail the epidemic. Public primary care facilities, where most people with HIV (PWH) in South Africa receive treatment, face myriad challenges retaining patients on antiretroviral therapy (ART). Nevertheless, some facilities manage to consistently retain PWH in care. We used a participatory Positive Deviance (PD) approach to discover characteristics of primary care facilities with above-average 12-month retention rates to develop an intervention. PD is an asset-based approach to behavior change that consists of discovering how high-performing outliers succeed despite sizable barriers, and then using those data to develop interventions for low-performers. Methods We conducted 11 in-depth leadership interviews, 9 staff focus groups with 29 participants, 11 patient focus groups with 45 participants, 23 patient shadowing visits, and 3 clinic observations in each of 3 high- and 3 low-retention public primary care facilities in Cape Town, South Africa, to discover characteristics of high-retention facilities that might be contributing to higher retention rates. Results Themes found to a greater degree in high-retention facilities were compassionate, respectful, patient-centered care; higher staff morale, passion for the work and team cohesion; efficient workflow procedures; and a welcoming physical environment. From these themes we developed the Connect intervention, consisting of strategies within three domains: (1) Engage, encourage, and support staff (e.g., a monthly staff support huddle, a compassion training); (2) Expedite and augment workflow procedures (e.g., adjust folder system to lower wait times); (3) Create a welcoming physical environment (e.g., fresh paint and plants in the waiting area). Conclusions A PD approach enabled us to identify factors that could be contributing to higher ART retention rates within low-resource public sector primary care facilities in Cape Town, South Africa. If effective, Connect could be a feasible, affordable complement to existing programmes aimed at improving care for PWH.

The use of positive deviance approach to improve health service delivery and quality of care: a scoping review

BACKGROUND

Quality has been a persistent challenge in the healthcare system, particularly in resource-limited settings. As a result, the utilization of innovative approaches is required to help countries in their efforts to enhance the quality of healthcare. The positive deviance (PD) approach is an innovative approach that can be utilized to improve healthcare quality. The approach assumes that solutions to problems are already available within the community and identifying and sharing those solutions can help others to resolve existing issues. Therefore, this scoping review aimed to synthesize the evidence regarding the use of the PD approach in healthcare system service delivery and quality improvement programs.

METHODS

Articles were retrieved from six international databases. The last date for article search was June 02, 2023, and no date restriction was applied. All articles were assessed for inclusion through a title and/or abstract read. Then, articles that passed the title and abstract review were screened by reading their full texts. In case of duplication, only the full-text published articles were retained. A descriptive mapping and evidence synthesis was done to present data with the guide of the Preferred Reporting Items for Systematic Reviews and Meta-analysis extension for Scoping Reviews checklist and the results are presented in text, table, and figure formats.

RESULTS

A total of 125 articles were included in this scoping review. More than half, 66 (52.8%), of the articles were from the United States, 11(8.8%) from multinational studies, 10 (8%) from Canada, 8 (6.4%) from the United Kingdom and the remaining, 30 (24%) are from other nations around the world. The scoping review indicates that several types of study designs can be applied in utilizing the PD approach for healthcare service and quality improvement programs. However, although validated performance measures are utilized to identify positive deviants (PDs) in many of the articles, some of the selection criteria utilized by authors lack clarity and are subject to potential bias. In addition, several limitations have been mentioned in the articles including issues in operationalizing PD, focus on leaders and senior managers and limited staff involvement, bias, lack of comparison, limited setting, and issues in generalizability/transferability of results from prospects perspective. Nevertheless, the limitations identified are potentially manageable and can be contextually resolved depending on the nature of the study. Furthermore, PD has been successfully employed in healthcare service and quality improvement programs including in increasing surgical care quality, hand hygiene practice, and reducing healthcare-associated infections.

CONCLUSION

The scoping review findings have indicated that healthcare systems have been able to enhance quality, reduce errors, and improve patient outcomes by identifying lessons from those who exhibit exceptional practices and implementing successful strategies in their practice. All the outcomes of PD-based research, however, are dependent on the first step of identifying true PDs. Hence, it is critical that PDs are identified using objective and validated measures of performance as failure to identify true PDs can subsequently lead to failure in identifying best practices for learning and dissemination to other contextually similar settings.

Cross-Sectional Analysis of Impulse Indicator Saturation Method for Outlier Detection Estimated via Regularization Techniques with Application of COVID-19 Data

Impulse indicator saturation is a popular method for outlier detection in time series modeling, which outperforms the least trimmed squares (LTS), M-estimator, and MM-estimator. However, using the IIS method for outlier detection in cross-sectional analysis has remained unexplored. In this paper, we probe the feasibility of the IIS method for cross-sectional data. Meanwhile, we are interested in forecasting performance and covariate selection in the presence of outliers. IIS method uses Autometrics techniques to estimate the covariates and outlier as the number of covariates P > n observations. Besides Autometrics, regularization techniques are a well-known method for covariate selection and forecasting in high-dimensional analysis. However, the efficiency of regularization techniques for the IIS method has remained unexplored. For this purpose, we explore the efficiency of regularization techniques for out-of-sample forecast in the presence of outliers with 6 and 4 standard deviations (SD) and orthogonal covariates. The simulation results indicate that SCAD and MCP outperform in forecasting and covariate selection with 4 SD (20% and 5% outliers) compared to Autometrics. However, LASSO and AdaLASSO select more covariates than SCAD and MCP and possess higher RMSE. Overall, regularization techniques possess the least RMSE than Autometrics, as Autometrics possesses the least average gauge at the cost of the least average potency. We use COVID-19 cross-sectional data collected from 1 July 2021 to 30 September 2021 for real data analysis. The SCAD and MCP select CRP level, gender, and other comorbidities as an important predictor of hospital stay with the least out-of-sample RMSE of 7.45 and 7.50, respectively.

Positive deviance in health and medical research on individual level outcomes - a review of methodology

BACKGROUND

Positive deviance as a methodology is increasing in application yet there is high variability in how this approach is applied in health services research.

METHODS

We conducted a scoping review of the literature for positive deviance applied to health outcomes informed by PRISMA-ScR. We searched the literature from 1945 to 2020, including articles on positive deviance or positive outliers, and restricted to examining individual rather than organizational outcomes. We analyzed the methodology applied including the process of identifying deviants, the use of control groups, and the degree of community engagement.

RESULTS

Our initial search identified 1140 manuscripts; we included 104 papers describing 98 studies, 11 topical and one miscellaneous category. Most studies used objective measures of health or survey-based responses to identify deviants from a sub-set of the population at risk. The use of controls was less common in some topics (hospital infections), whereas controls were universally applied in other topics (malnutrition). The degree of community engagement varied widely.

CONCLUSIONS

Positive deviance would benefit from improvements in reporting and standardized approaches to defining deviance. Studies could be improved through clarified definitions of deviance/risk, explicit descriptions of community engagement, and more consistent use of controls.

A clustering approach for detecting implausible observation values in electronic health records data

Background

Identifying implausible clinical observations (e.g., laboratory test and vital sign values) in Electronic Health Record (EHR) data using rule-based procedures is challenging. Anomaly/outlier detection methods can be applied as an alternative algorithmic approach to flagging such implausible values in EHRs.

Methods

The primary objectives of this research were to develop and test an unsupervised clustering-based anomaly/outlier detection approach for detecting implausible observations in EHR data as an alternative algorithmic solution to the existing procedures. Our approach is built upon two underlying hypotheses that, (i) when there are large number of observations, implausible records should be sparse, and therefore (ii) if these data are clustered properly, clusters with sparse populations should represent implausible observations. To test these hypotheses, we applied an unsupervised clustering algorithm to EHR observation data on 50 laboratory tests from Partners HealthCare. We tested different specifications of the clustering approach and computed confusion matrix indices against a set of silver-standard plausibility thresholds. We compared the results from the proposed approach with conventional anomaly detection (CAD) approaches, including standard deviation and Mahalanobis distance.

Results

We found that the clustering approach produced results with exceptional specificity and high sensitivity. Compared with the conventional anomaly detection approaches, our proposed clustering approach resulted in significantly smaller number of false positive cases.

Conclusion

Our contributions include (i) a clustering approach for identifying implausible EHR observations, (ii) evidence that implausible observations are sparse in EHR laboratory test results, (iii) a parallel implementation of the clustering approach on i2b2 star schema, and (3) a set of silver-standard plausibility thresholds for 50 laboratory tests that can be used in other studies for validation. The proposed algorithmic solution can augment human decisions to improve data quality. Therefore, a workflow is needed to complement the algorithm’s job and initiate necessary actions that need to be taken in order to improve the quality of data.

Electronic supplementary material

The online version of this article (10.1186/s12911-019-0852-6) contains supplementary material, which is available to authorized users.